Background Eating disorders rank among the ten leading causes of disability among young women, and anorexia nervosa has the highest mortality rate of all mental disorders. Follow-up studies have shown that 20-30 % of patients with eating disorders develop longstanding symptoms, seriously impairing their daily and represents a public health concern. There are very few studies on the course of these patients. Several studies have demonstrated comorbidity between eating disorders and personality disorders. Among patients with eating disorders, the reported frequencies of personality disorders vary from 27% to 77%. Most of the studies are cross-sectional designs, thus unable to catch trends or changes over time. There is a need for prospective longitudinal studies of adult patients using structured diagnostic interviews both for eating disorders and personality disorders.

At Modum Bad, a Norwegian psychiatric hospital, the investigators have conducted a follow-up study of patients with longstanding eating disorder 1-, 2- and 5-years after treatment. The aim of the present project is to follow-up the patients additional 17-years after treatment.

Objective Investigate the 17-years course and outcome of adult patients with severe and longstanding eating disorders with regard to eating disorder-related symptoms, general symptoms and personality disorders in addition to examining whether personality disorders and sexual abuse in childhood can predict the course and outcome.

Method Examining patients 17-years after treatment with standardized interviews and questionnaires.

Participants From August 1998 to June 2001, 92 consecutive patients were admitted to a specialized eating disorder unit at a psychiatric hospital, Modum Bad. The admission criteria were symptoms of anorexia nervosa, bulimia nervosa, and eating disorder not otherwise specified that impaired daily life functioning, inadequate responses to previous treatment and age above 18 years. Those with severe medical complications and body mass index (BMI) < 14 were excluded. Patients were allocated to specific inpatient treatment programs for anorexia nervosa and bulimia nervosa respectively. Six patients discharged themselves during the first two weeks. A total of 86 remaining patients, one male and 85 female, initially entered the study. Participants have been assessed at five points in time, i.e. 1 (Admission, 1998-2001), 2 (Discharge, 1998-2001), 3 (one year after admission, 2001-2002), 4 (two years after admission, 2002-2003) and time 5 (five years after admission, 2004-2005). A total of 77 patients (90 %) were available for the five-year follow-up. These patients have given written consent to participate in a 17-year follow-up. Attrition is a common problem in long-term follow-up investigations, but appears to be less severe in Norway generally and at Modum Bad in particular.

Aims Study 1: Investigate the course and outcome of eating disorder patients Subordinate aims included: to a) report on the 17-year course and outcome of adults with longstanding eating disorder and b) identify eating disorder subgroups with various courses and outcomes.

Study 2: Investigate the course of personlaity disorders among eating disorder patients Subordinate aims included: to a) report the presence of categorical and dimensional personality disorder in adults with longstanding eating disorder at admission to hospital treatment and at 1-, 2- 5-year and 17-year follow-up; b) compare the changes of personality disorder in anorexia nervosa, bulimia nervosa, and eating disorder not otherwise specified; c) investigate if personality disorder is less frequent among recovered patients; d) compare the magnitude of changes in specific eating disorder symptoms, general psychopathology and personality disorder; e) investigate whether changes in personality disorder influence changes in symptom scores, or vice versa.

Study 3: Identify predictors of poor outcome in longstanding eating disorder Subordinated aims included: to a) investigate if personality disorder and child sexual abuse predict the course of severity of eating disorder symptoms, b) examine how child sexual abuse and personality disorder interact to affect the eating disorder course. Does persoanlity disorder mediate, does child sexual abuse moderate, or do they independently influence the eating disorder course?

Feasibility: design, method, analysis The present sample is unique in having participating in a previous prospective longitudinal study of adult patients using structured diagnostic interviews both for the eating disorder and the personality disorder. The very longterm course and outcome can be evaluated. There are no other studies where several repeated structured clinical interviews have been used to assess personality disorders in a sample of patients with eating disorders. This study may bridge the gap between science and practice in the assessment and treatment of severe eating disorder patients. As there are no generally accepted guidelines for treatment of severe eating disorder, the present finding may elicit efforts to develop such guidelines. The main finding in the previous studies is that a majority of patients with severe and longstanding eating disorder experience significant improvement that is fairly stable through years. This knowledge stands in contrast to a widespread professional opinion that there is little chance of improvement for individuals with longstanding eating disorder. Bridging the gap between science and clinical practice the refuting of this opinion by dissemination of research findings stands out as an important implication as one aspect of being a clinician is to provide a realistic hope to patients and relatives. However the previous studies do not provide the full knowledge of the full course of eating disorder. Thus, a further extension of the study is important.

Inclusion Criteria:

• Symptoms of anorexia nervosa, bulimia nervosa, and eating disorder not otherwise specified that impaired daily functioning
• Inadequate responses to previous treatment
• Age above 18

Exclusion Criteria:

• Serious medical complication
• Body mass index (BMI) below 14

Adult patients with longstanding eating disorder